Re: [Beowulf] Re: Beowulf Digest, Vol 46, Issue 12

I am new to the parallel computing but have the desire to learn more and to utilize this technology for my organization needs. Can any body guide me for better available resources? parallel computing has been around a long time; lots of good books out there. but what you should first do is look

[Beowulf] Re: Beowulf Digest, Vol 46, Issue 12

I am new to the parallel computing but have the desire to learn more and to utilize this technology for my organization needs. Can any body guide me for better available resources? Ali Zahoor ___ Beowulf mailing list, Beowulf@beowulf.org To change you

Re: [Beowulf] multi-threading vs. MPI

Thanks for the results, and the link. In section 6.7 of the NAS Parallel Benchmark (NPB 2.1 Results Report, NAS-95-010 (PDF-213KB) on MPI, I found a discussion of the Clustered-SMP issues discussed so far in this thread. Its i

Re: [Beowulf] multi-threading vs. MPI

Greg Lindahl wrote: On Tue, Dec 11, 2007 at 08:52:11AM -0500, Joe Landman wrote: On the contrary, it is precisely because people are asking "how should I parallelize" that they need to ask the basic question of "where does my code spend time for my problems." OK, so say I have a garden-varie

Re: [Beowulf] multi-threading vs. MPI

On Tue, Dec 11, 2007 at 08:52:11AM -0500, Joe Landman wrote: > On the contrary, it is precisely because people are asking "how should I > parallelize" that they need to ask the basic question of "where does my > code spend time for my problems." OK, so say I have a garden-variety finite-differe

Re: [Beowulf] ever heard of ScaleMP?

[EMAIL PROTECTED] wrote: Abaqus is now MPI capable (the first of the implicit FEM codes that I know of). I'm happy to correct you that Actran (http://www.fft.be/?id=10) is already MPI capable since 2003. Actran is is also an implicit FEM code but focused on aero-acoustics and vibro-acoustics

Re: [Beowulf] ever heard of ScaleMP?

On Tue, Dec 11, 2007 at 09:28:17AM -0500, Mark Hahn wrote: > there's a company, ScaleMP, which seems to be selling some kind of > kit which enables to fairly large shared-memory x86_64 systems. As far as I can tell, they are just another software distributed shared memory company. Which has been

Re: [Beowulf] multi-threading vs. MPI

At 05:17 AM 12/11/2007, Douglas Eadline wrote: This is indeed the issue. Where to invest time? My opinion, and it is only my opinion, is the following. Please share your own. Threaded approaches do not scale across clusters. The memory architecture of multi-core is making nodes look more like s

Re: [Beowulf] ever heard of ScaleMP?

On Tue, 2007-12-11 at 09:28 -0500, Mark Hahn wrote: > there's a company, ScaleMP, which seems to be selling some kind of > kit which enables to fairly large shared-memory x86_64 systems. > their website is nearly useless (http://www.scalemp.com/), but a little > more info can be had from SGI, whi

Re: [Beowulf] ever heard of ScaleMP?

On Tue, 11 Dec 2007 at 7:19am, [EMAIL PROTECTED] wrote Abaqus is now MPI capable (the first of the implicit FEM codes that I know of). So ScaleMP isn't needed for the newer version of Abaqus. The implicit mode of LS-DYNA is also MPI capable, as of ls971 (i.e., a year or so ago, IIRC). -- Jo

Re: [Beowulf] ever heard of ScaleMP?

I've been quite curious to try something like the f1200 as a potential replacement for our Altixes, which were bought predominantly for running single-threaded large-memory jobs. we have an Altix as well, and I always cringe when I see a single-thread, large-memory job running on it. ours has

Re: [Beowulf] ever heard of ScaleMP?

On Tue, Dec 11, 2007 at 07:19:07AM -0800, [EMAIL PROTECTED] wrote: > Mark Hahn <[EMAIL PROTECTED]> wrote: > > > > SGI seems to aim it purely at structural/cfd/crash sims - > > mainly using Abaqus and related tools. > > Abaqus is now MPI capable (the first of the implicit FEM codes that I k

Re: [Beowulf] ever heard of ScaleMP?

Flextronics was showing a small cluster where they had 4 boxes connected by IB and within each box they had 4 systems connected by IB. They were running ScaleMP on it. They had a graph of running Stream on top of the system. They were plotting bandwidth vs. number of cores and it was fairly linear

Re: [Beowulf] ever heard of ScaleMP?

On 11 Dec 2007, at 4:53 pm, Joe Landman wrote: Tim Cutts wrote: I've been quite curious to try something like the f1200 as a potential replacement for our Altixes, which were bought predominantly for running single-threaded large-memory jobs. It is fairly easy (barring cost issues) to ge

Re: [Beowulf] ever heard of ScaleMP?

Tim Cutts wrote: I've been quite curious to try something like the f1200 as a potential replacement for our Altixes, which were bought predominantly for running single-threaded large-memory jobs. It is fairly easy (barring cost issues) to get a single system image machine with 8-16 processor

Re: [Beowulf] ever heard of ScaleMP?

On 11 Dec 2007, at 3:19 pm, <[EMAIL PROTECTED]> wrote: Mark Hahn <[EMAIL PROTECTED]> wrote: there's a company, ScaleMP, which seems to be selling some kind of kit which enables to fairly large shared-memory x86_64 systems. their website is nearly useless (http://www.scalemp.com/), but a

Re: [Beowulf] WSEAS, Call for Papers

I just recreated an article "WSEAS" at the English language Wikipedia, as I had had to hunt around some to find out what WSEAS is. It turns out that there had been an aritlce previously, but it was deleted for "copyright violation". Almost always that means someone from an organization created an a

Re: [Beowulf] multi-threading vs. MPI

Geoff Jacobs wrote: Please note that we are all speaking as developers on clusters. Although this is a valuable niche in the market, it is still only a niche. Even with workstations, systems still only come with a few cores. In this regime, threads are still relevant from a performance standpoin

RE: [Beowulf] multi-threading vs. MPI

I'm no expert in CPMD but - AFAIK CPMD is not meant to run as pure OpenMP; OpenMP parallelization is auxiliary to MPI - in the paper " Dual-level parallelism for ab initio molecular dynamics: Reaching teraflop performance with the CPMD code" the developers (Jürg Hutter and Alessandro Curioni) c

Re: [Beowulf] ever heard of ScaleMP?

Joe Landman <[EMAIL PROTECTED]> wrote: > Mark Hahn wrote: > > there's a company, ScaleMP, which seems to be selling some kind of kit > > which enables to fairly large shared-memory x86_64 systems. > > their website is nearly useless (http://www.scalemp.com/), but a little > > more info can b

Re: [Beowulf] ever heard of ScaleMP?

Mark Hahn <[EMAIL PROTECTED]> wrote: > there's a company, ScaleMP, which seems to be selling some kind of > kit which enables to fairly large shared-memory x86_64 systems. > their website is nearly useless (http://www.scalemp.com/), but a little > more info can be had from SGI, which apparen

Re: [Beowulf] multi-threading vs. MPI

Douglas Eadline wrote: > This is indeed the issue. Where to invest time? > > My opinion, and it is only my opinion, is the following. > Please share your own. > > Threaded approaches do not scale across clusters. The memory > architecture of multi-core is making nodes look more like > small clust

Re: [Beowulf] ever heard of ScaleMP?

Mark Hahn wrote: there's a company, ScaleMP, which seems to be selling some kind of kit which enables to fairly large shared-memory x86_64 systems. their website is nearly useless (http://www.scalemp.com/), but a little more info can be had from SGI, which apparently uses ScaleMP for their f1200

[Beowulf] ever heard of ScaleMP?

there's a company, ScaleMP, which seems to be selling some kind of kit which enables to fairly large shared-memory x86_64 systems. their website is nearly useless (http://www.scalemp.com/), but a little more info can be had from SGI, which apparently uses ScaleMP for their f1200 product (rebadged

Re: [Beowulf] multi-threading vs. MPI

Greg Lindahl wrote: On Mon, Dec 10, 2007 at 07:40:15PM -0500, Joe Landman wrote: b) the only code that matters is your code (really, this should be everyone's mantra with benchmarking in general). Well, the problem here is that the question most people are asking is, "how should I parallelize

Re: [Beowulf] multi-threading vs. MPI

This is indeed the issue. Where to invest time? My opinion, and it is only my opinion, is the following. Please share your own. Threaded approaches do not scale across clusters. The memory architecture of multi-core is making nodes look more like small clusters i.e. memory is becoming more locali